The associations were strongest for beer and liquor, intermediate for white wine, and weakest for red wine. Multivariate adjustment weakened the association of myocardial infarction with wine consumption but strengthened the associations with beer and liquor consumption.